Parallel Sets Transition Splice

[bwat]

Parallel sets of 750s, but the lugs at the end only handle max 500s. It was 750kcmil for voltage drop. Contractor wants to splice each 750 individually to a 500 just a few feet from termination point.

I interpret 310.10(H) to say that this is not allowed and that all the same phase conductors need to be tied together when terminating and splicing, even for something like this, where they will be tied together just a few feet later.

Am I correct?

The length of the 500s is so small in comparison to the length of the 750s, so I can't see it creating any real world problems, but you know... code is code.

 

[dkidd]

Parallel sets of 750s, but the lugs at the end only handle max 500s. It was 750kcmil for voltage drop. Contractor wants to splice each 750 individually to a 500 just a few feet from termination point.

I interpret 310.10(H) to say that this is not allowed and that all the same phase conductors need to be tied together when terminating and splicing, even for something like this, where they will be tied together just a few feet later.

Am I correct?

The length of the 500s is so small in comparison to the length of the 750s, so I can't see it creating any real world problems, but you know... code is code.

You are correct.

 

[bwat]

You are correct.

Thank you

 

[augie47]

I'm sure I'm overlooking something but if each conductor is treated the same and the ampacity of the 500s is adequate, I can't see the violation. Very unusual but violation ?
Have you looked at using pin sleeves on on the 750 ? All this may be a moot point using them.

 

[Carultch]

You are correct.

I don't agree that the NEC requires you to bring them all together, just to step down the size for adapting to a termination. For tapping, it is another matter entirely, because you introduce an imbalance in current if you inconsistently tap the sets.

Solely for splicing, as long as you make each path in parallel as identical as possible per what the NEC requires, you've met the intent of the code. This would mean using identical splices, and identical lengths/sizes/types for the section of conductor that terminates in the equipment.

 

[dkidd]

(1) General. Aluminum, copper-clad aluminum, or copper
conductors, for each phase, polarity, neutral, or grounded
circuit shall be permitted to be connected in parallel (electrically
joined at both ends) only in sizes 1/0 AWG and
larger where installed in accordance with 310.10(H)(2)
through (H)(6).

 

[bwat]

The fact that there's even a mild debate on this affirms my uncertainty around this.

Would this underlined part (underlined by me) be applicable? I think you could make an argument that they aren't doing this if they step down to a different wire.

310.10(H)(1)
Aluminum, copper-clad aluminum, or copper
conductors, for each phase, polarity, neutral, or grounded
circuit shall be permitted to be connected in parallel (electrically
joined at both ends) only in sizes 1/0 AWG and larger
where installed in accordance with 310.10(H)(2) through
(H)(6).

Edit: ha! I see dkidd beat me by a few seconds.

 

[dkidd]

See recent thread:

Polaris Taps in MDP

Mike this is Bryan Schultz with C. Davis Electric. Ed Davis wanted me to send you a message about an issue that come up today with an inspector. It is regarding polaris taps in an MDP. We have a 600 amp 600v breaker with 4 sets of 350 mcm. We had to polaris tap these 4 sets with 2 sets of 350 to...

forums.mikeholt.com

 

[david luchini]

The fact that there's even a mild debate on this affirms my uncertainty around this.

Would this underlined part (underlined by me) be applicable? I think you could make an argument that they aren't doing this if they step down to a different wire.


Edit: ha! I see dkidd beat me by a few seconds.

The parallel sets are still electrically joined at the end, after the splice.

 

[bwat]

The parallel sets are still electrically joined at the end, after the splice.

After the splice is not the end of that first parallel set. At the splice you're starting a different parallel set.

 

[Strathead]

The parallel sets are still electrically joined at the end, after the splice.

I agree. One other thing that beers noting. The 500KCMIL is capable of carrying the circuit amps so the 750 was upsized for voltage drop or other.

 

[LarryFine]

After the splice is not the end of that first parallel set. At the splice you're starting a different parallel set.

I disagree. As long as each conductor is spliced the same way, the paralleled conductors maintain the same characteristics.

 

[infinity]

The parallel sets are still electrically joined at the end, after the splice.

I agree, this is a fairly common installation practice and is code complaint.

 

[synchro]

Splicing the short 500s to the 750s individually and then connecting the 500s together should, if anything, improve the balance of the currents through the 500s and into the termination. That's because the larger resistance of the long 750s will dominate over the small resistances of the individual termination connections and therefore establish how well the currents are shared, even if these small termination resistances have mismatch.

 

[Carultch]

Splicing the short 500s to the 750s individually and then connecting the 500s together should, if anything, improve the balance of the currents through the 500s and into the termination. That's because the larger resistance of the long 750s will dominate over the small resistances of the individual termination connections and therefore establish how well the currents are shared, even if these small termination resistances have mismatch.

That's a good point too.

If anything, requiring you to make perform your splice where you bring all wires together, is counterproductive to the intent of the context rules of parallel conductors.

The language of "joined at both ends" in the beginning of the 310.10(H) needs to use that language, for defining what it means to use conductors in parallel in the first place, but there is no physical basis for not allowing splicing/reducing within an individual set.. That is, as long as all the remaining sets also get spliced and reduced in the same manner. If that were the case, then every time you make a splice even between identical sized conductors, you'd need to rejoin them in parallel.

To give an example, suppose you had about 100 ft of 750's, and about 1 foot of 500's at the termination.

We know that there is going to be some practical amount of inconsistency in length on both of these dimensions, even if the code gives no margin of error to address how much error is allowable. Suppose the two lengths of 750's were 101 ft and 100 ft, and that the two lengths of 500's were 13" and 12". That takes a disciplined amount of skill to get it that close. Going by the approximation that resistance is proportional to length/kcmil, the following is how it divides:

Combining the 750's in this example on their own, results in a 50.25% / 49.75% division of the current.
Combining the short lengths of 500's on their own, results in a 52% / 48% division of the current.
Splice together the sizes without paralleling, and when the two longest lengths are in series, you get a 50.3%/49.7% division of the current.

So you can see that the 1" of difference for the short whip at the termination is a lot more significant than the 1 foot of difference for the majority of the run. By requiring you to parallel prior to splicing, you end up with a greater split between the way the current divides between the two parallel sets.